mysql主从邮件监控
2016-09-06 本文已影响93人
syncwt
概述
包括mysql宕机报警,mysql主从io,sql状态异常报警以及主从延迟过长报警
通过外部STMP发邮件
# vi /etc/mail.rc
添加以下设置:
# 发送邮件邮箱
set from=WENTAO_Wanna@126.com
# 外部stmp服务器地址
set smtp=smtp.126.com
# 外部smtp服务器认证的用户名
set smtp-auth-user=WENTAO_Wanna
# 外部smtp服务器认证的用户密码,注意是外部服务器的授权码并非邮箱登录密码
set smtp-auth-password=123456
# 邮件认证的方式
set smtp-auth=login
调试
# echo mail_content | mail -s "mail_title" WENTAO_Wanna@126.com,WENTAO_Wanna@foxmail.com
shell脚本实现自动警告邮件提示
-
创建保存警告文件的路径:
/data/mysql_warning
# cd /data # mkdir mysql_warning
-
创建备份脚本文件:
/data/mysql_bak/mysql_monitoring.sh
# cd /data/mysql_warning # touch mysql_monitoring.sh # vi mysql_monitoring.sh
输入以下内容:
#check MySQL Slave's Runnning Status #Crontab time 00:10 MYSQLPORT=`netstat -na|grep "LISTEN"|grep "3306"|awk -F[:" "]+ '{print $4}'` MYSQLIP=`ifconfig eth0|grep "inet" | awk ' {print $2}'` STATUS=$(/usr/bin/mysql -uroot -pyour-password -S /data/mysql/mysql.sock -e "show slave status\G" | grep -i "running") DELAYED=$(/usr/bin/mysql -uroot -pyour-password -S /data/mysql/mysql.sock -e "show slave status\G" | grep -i "Seconds_Behind_Master") #echo "$STATUS" IO_env=`echo $STATUS | grep IO | awk ' {print $2}'` echo "=============================" echo "$STATUS" echo "$DELAYED" SQL_env=`echo $STATUS | grep SQL | awk ' {print $4}'` DELAYED_env=`echo $DELAYED | awk ' {print $2}'` DATA=`date +"%y-%m-%d %H:%M:%S"` function checkMysqlStatus(){ if [ "$MYSQLPORT" == "3306" ] then /usr/bin/mysql -uroot -pyour-password --connect_timeout=5 -e "show databases;" &>/dev/null 2>&1 if [ $? -ne 0 ] then echo "Server: $MYSQLIP mysql is down, please try to restart mysql by manual!" > /data/mysql/mysql.err mail -s "WARN! server: $MYSQLIP mysql is down." WENTAO_Wanna@126.com < /data/mysql/mysql.err else echo "mysql is running..." fi else mail -s "WARN!Server: $MYSQLIP mysql is down." WENTAO_Wanna@126.com fi } echo "=================================" echo "$IO_env" echo "$SQL_env" echo "$DELAYED_env" echo "=================================" checkMysqlStatus # 延时大于60s if [ "$DELAYED_env" -ge 60 ] then echo "MySQL Slave is delayed $DELAYED_env s!" echo "####### $DATA #########">> /data/mysql_slave/mysql_slave_status.log echo "MySQL Slave is delayed $DELAYED_env s!" >> /data/mysql_slave/mysql_slave_status.log echo "MySQL Slave is delayed $DELAYED_env s!" | mail -s "WARN! $MYSQLIP MySQL Slave is delayed $DELAYED_env s!" WENTAO_Wanna@126.com fi if [ "$IO_env" = "Yes" -a "$SQL_env" = "Yes" ] then echo "MySQL Slave is running!" else echo "####### $DATA #########">> /data/mysql_slave/mysql_slave_status.log echo "MySQL Slave is not running!" >> /data/mysql_slave/mysql_slave_status.log echo "MySQL Slave is not running!" | mail -s "WARN! $MYSQLIP MySQL Slave is not running." WENTAO_Wanna@126.com fi
检测参数说明:
- Slave_IO_Running:该参数可作为io_thread的监控项,Yes表示io_thread的和主库连接正常并能实施复制工作,No则说明与主库通讯异常,多数情况是由主从间网络引起的问题;
- Slave_SQL_Running:该参数代表sql_thread是否正常,具体就是语句是否执行通过,常会遇到主键重复或是某个表不存在。
- Seconds_Behind_Master:是通过比较sql_thread执行的event的timestamp和io_thread复制好的event的timestamp(简写为ts)进行比较,而得到的这么一个差值;
NULL—表示io_thread或是sql_thread有任何一个发生故障,也就是该线程的Running状态是No,而非Yes。
0 — 该值为零,是我们极为渴望看到的情况,表示主从复制良好,可以认为lag不存在。
正值 — 表示主从已经出现延时,数字越大表示从库落后主库越多。
负值 — 几乎很少见,我只是听一些资深的DBA说见过,其实,这是一个BUG值,该参数是不支持负值的,也就是不应该出现。
-
修改文件属性,使其可执行
# chmod +x /data/mysql_warning/mysql_monitoring.sh
-
修改
/etc/crontab
# vi /etc/crontab
在下面添加
10 00 * * * root /data/mysql_warning/mysql_monitoring.sh
表示每天早上8点至晚上12点每30分钟执行一次监控脚本
-
重新启动
crond
使设置生效# systemctl enable crond.service #设为开机启动 # systemctl restart crond.service #重新启动
-
手动执行效果
# sh /data/mysql_warning/mysql_monitoring.sh
Warning: Using a password on the command line interface can be insecure.
Warning: Using a password on the command line interface can be insecure.
=============================
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
Slave_SQL_Running_State: Slave has read all relay log; waiting for the slave I/O thread to update it
Seconds_Behind_Master: 0
=================================
Yes
Yes
0
=================================
mysql is running...
MySQL Slave is running!
邮件效果
邮件效果图参考
LINUX下通过外部SMTP发邮件(直接抛弃sendmail和postfix)
Seconds_Behind_Master分析主从延迟